Hammering: Gaining and Maintaining Control in Dexterous Piano Playing

When students play piano passages that push the student to (or near) their limit of dexterity, every student except the extremely gifted will demonstrate a lack of control. Such passages could consist of several lines or pages of quick-running notes such as Burgmuller’s “Velocity,” or it could just be quick notes dispersed sporadically through a piece such as Ellmenreich’s “Spinning Song” (especially when the left hand takes over the melody in the middle section).   I define this “lack of control” to be one or more of the following:

  • Rhythmic unevenness
  • Dynamic unevenness (some notes too soft or ghosted altogether, and/or various unintended accents)
  • Lack of clarity (no precision in lifting fingers)
  • Lack of togetherness between hands1

This control problem arises when kinesthetic memory of rapid muscle movements starts to become so fluid that the student begins to forget what it was like to perceive every note in the piece with individual clarity. For example, instead of the brain giving five separate commands to the fingers to play five separate notes, the brain starts to mush all five fingers together into one single command, similar to how we might be tempted to slur words together in a phrase after we’ve spoken the phrase a thousand times. Consequently, certain fingers play earlier than they should2, softer than they should, they don’t remove from the keys quickly enough, and there is less control over synchronization between the hands. When a student cognitively understands what their fingers are supposed to be doing, but their fingers don’t obey in a manner that sounds controlled, I call this muscu-blur (muscular blur).

A pianist is forced to experience piano playing with a feeling of "muscular blur."
Shown above: the imagined hand/finger configuration whenever music sounds like it is in desperate need of hammering.

This is why we notice our students’ pieces will most naturally deteriorate in quality week by week after they “finished” their piece. The greatest pianists never stop perceiving the individuality of every note they play, even when they’re playing 10 or more notes per second, even when they finished the piece five years ago. But students who perform their difficult piece over and over again and never actually practice the piece slowly are doomed to muscu-blur.

Most teachers have certain practice recommendations for students to address each of these issues. For example, I think most of us are familiar with the routine of playing long-short (swinging) pairs followed by short-long (inverted swinging) pairs in order to correct rhythmic unevenness. Dynamic unevenness can be corrected by playing the first of every four notes accented (or in compound meter, every three notes), followed by playing the second note accented, then the third, and so on. These approaches do work since they force the student to pay special, individualized attention to every single note. Unfortunately, these practice techniques are inefficient. Just one iteration of the rhythmic exercise requires playing a passage (or entire piece) twice, while the dynamic accent exercise requires playing it three or four times. There is also a certain tempo limit that cannot be surpassed with these exercises – one can only swing or accent notes at a very slow tempo (it cannot be done while playing “ripping fast” passagework unless it is slowed down drastically). Why not instead use a practice technique that only requires one iteration and does not have any tempo ceiling? Better yet, why not use a practice technique that can address all four of the above technical weaknesses simultaneously with one iteration?

That is where a practice technique I call “hammering” comes in. My definition of hammering is when one practices at the fastest tempo one can control with absolute perfection (which is always slower than the performance speed when the piece isn’t done yet), playing every note in both hands forte, clearly, rhythmically even, hands perfectly together, all with relaxed arms/hands and no pedaling. All dynamics, balance and voicing are ignored (both hands play equally loud). If it is done right, it makes it sound like a computer is playing (or perhaps like the pianist is angry at the piano, although forte suffices – playing fortissimo is unnecessary).

Origin And Reinforcement

I first learned this technique as an undergraduate from Mark Westcott3 (1967 Van Cliburn winner), when he told me to play the difficult middle section passage of Chopin’s B-flat minor Scherzo over and over again with “all notes loud.” This was after I complained that no matter how much I practiced that nasty section, it simply did not get any better. Hammering produced instant results. I started utilizing the technique whenever there was any lack of control in my playing in passagework demanding dexterity.  It had effects that were nothing short of phenomenal.

A few years later, a college classmate was learning the Chopin F minor Ballade, and he came to me one day and said he reached a ceiling – no matter how much he practiced, it just wouldn’t get any better. This sounded <cough> strangely familiar to me. Like a doctor prescribing a pill, I told him to go hammer these difficult passages several times each day for the next couple weeks. The next time I saw him, he said it was the single most useful advice anyone had ever given him to help him play piano.  He later played the piece brilliantly in his undergraduate recital.

Years later, Jon Nakamatsu (1997 Van Cliburn Competition gold medalist) came to my local city. While he was here, he gave a master class. Two of the three students playing in that master class had studied with me for 5-6 years, roughly up to the level of being able to play Schubert Impromptus and certain movements of Mozart and Beethoven sonatas. When these students left my studio (disgruntled parents due to clashes with my policy in my dark days of having a rigid make-up lesson policy), they were both “distinction” students, meaning they achieved a highly-coveted “distinction” rating in our local festival each year, a rating that had to be unanimous among all three judges (a rating that listed among its criteria, “a knock-your-socks-off performance”). Typically, only 4-6 students each year out of around 100 participating students would achieve this rating. Both students were hammering experts. They both hammered only those passages that needed hammering, only as much as needed, and I never had to talk about hammering with them – they were at a point where using it was instinctual on their part. Their playing showed it, because it was always so beautifully controlled.

Within three measures of hearing these students play the faster sections of their music, I could tell that they had been untrained to hammer by their new teacher. It was painfully obvious. I knew these students well enough to know that they would continue hammering forever unless told otherwise. They had to have been told otherwise. (We’ll talk later about why some teachers would want to reject the single most helpful practice technique known to the world of pianism.) When the first student finished, the very first thing Jon said (after complimenting the student on various aspects of musicality, of course) was that they needed to “be able to play their pieces like a computer.” Those were his exact words. He explained that he should be able to play all of his uncontrolled passagework first in such a way where the volume of every note is normalized to a loud volume, before they can hope to control the passagework any other way.  When the second student finished, he said the exact same thing.

Notice my emphasis of “be able to” in the Jon Nakamatsu quote above. This practice routine has nothing to do with whether we like how hammering sounds or whether such playing would want to be heard on stage. Obviously, it sounds terrible. But if one is incapable of playing a piece like a computer, how in the world would they be capable of playing it with perfect musical inflection and control? I myself can’t help gravitating to the same computer analogy with my students to describe hammering to them since I’ve had so much experience hearing how computers perform music – as in, what you hear after you input notes into any music notation program using a mouse and have the computer play it back.4

I train all of my students to hammer, beginning with the first time control becomes an issue in a passage requiring dexterity (usually at the intermediate level), and I regularly get similar voluntary, unprompted feedback from them about how effective it is. Many practice techniques are effective, but none seem to produce as dramatic and sweeping results as hammering. It truly is a magic bullet.

How It Works

Some say hammering is “unmusical.”  It depends on what definition of “musical” we are talking about.  If we’re talking about, “Something that sounds musically pleasing,” then they’re right: this is unmusical. But if we’re talking about, “Something that requires great musicianship to accomplish” (the definition I have in mind), then they’re wrong. It takes every bit as much dynamic discernment and concentration to detect a dynamic flaw that deviates by only 10% in a series of notes played equally loud as it does to detect a dynamic flaw in a series of notes played as a crescendo or diminuendo. It requires intense musical listening. And we’re not just listening for dynamic evenness when we hammer: we’re listening for rhythmic evenness, togetherness and clarity as well. I have spent small portions of lessons for many weeks with some students trying to get them to hammer in just the right way. When they find it, control over their playing skyrockets.

Let’s now compare the likelihood of musical success for hammering vs. accenting every third or fourth note.  Supposing 10 is fortissimo while 1 is pianissimo, and a student tries to play notes with the following dynamic pattern:

10  1  1  1  10  1  1  1

…but the student plays instead:

8  2  1  3  9  2  2  1

Inevitably the student concludes their mission has been accomplished, because it’s so difficult to compare the 8’s and 9’s together when they don’t occur one after another. The 1s, 2s and 3s are also difficult to judge against each other because of the surrounding (distracting) accents.  This method of addressing dynamic evenness is somewhat successful but still leaves holes.

But now suppose the student tries to play the following dynamic pattern:

8   8   8   8   8   8   8   8……

Every time the student plays a 7 or 9, these minute differences will be easier to detect because we are now comparing apples right next to each other, instead of comparing apples with three oranges between.

As if killing four birds (rhythmic evenness, dynamic evenness, clarity, togetherness) with one stone weren’t enough, here are three more birds for you:

1. Reinforcement of kinesthetic memory:  When every note is played loudly, it gives the pianist greater kinesthetic sensation both in movement and in feedback for every note played. Norman Doidge (p. 63-64) discusses research that M. Merzenich designed or helped carry out over the past few decades, studying the brain maps of monkeys, finding that after months of living with two fingers that are sewn together, a monkey’s brain would map those two fingers as one. Any stimulus on either of the two sewn fingers would light up the same part of the brain map. These results were also confirmed on humans with “webbed-finger syndrome,” and surgically separating the two fingers caused the human brain to remap each finger individually. This demonstrates that timing of sensations in the fingers affect the brain’s map of the fingers: when two fingers always experience the same sensations at the same time, those two fingers are eventually treated as one, and when the brain perceives independence in finger sensations (both in movement and in feedback), the brain’s map of those fingers is refined for greater independence and control. Neuroscience now knows this as Neurons that fire apart wire apart – or Neurons out of sync fail to link. Hammering is done at a slightly slower tempo than the performance tempo, allowing the pianist enough time to redevelop their sense of individuality of each note, and when those sensations are made more intense (loud playing), the return of this feeling is hastened.

2. Reinforcement of audio memory:  The feedback we get from sound works the same way as tactile feedback does. We pianists tend to remember notes (sound) with more clarity than we remember rests (lack of sound). Hearing every note played with equal dynamic importance reminds us of all the notes that should be executed with equal conscious attention during performance, regardless of expression.

3. Reinforcement of analytical memory:  Because it involves practicing slower so that every note is totally controlled, it ensures that one is still mentally engaged on a microscopic level when playing at a fast speed.  Fast tempos have a way of making us forget about the nitty-gritty details we had digested when we first learned the piece, and so memory weaknesses develop. I think hammering is one of the reasons I’ve always recovered from wrong notes and slips quickly. Mistakes I’ve made are almost never a function of not remembering what I’m supposed to play – it’s usually because I physically get “off track” then must get back on track, mishitting notes or slipping off of keys.  This isn’t to say that it makes students analyze music correctly or thoroughly – it’s just to say that the student will hold on to as much of their own level of analysis as is practical for them.

I know every teacher has both experienced in their own playing and experienced in their students’ playing the phenomenon of a finished piece deteriorating over time. This goes back to what I described above, the lack of individual perception of every note. Deteriorating pieces can only be fixed by reinforcing the individuality of every note. Nothing is more efficient at doing this than hammering.

I call it hammering, Nakamatsu calls it playing like a computer, Westcott calls it playing every note loud. Whatever you call it, it works. In my book, a fast piece has no hope of being considered “done” technically if one is incapable of hammering it perfectly all the way through – it will still demonstrate technical holes.

Pitfalls

If students come to lessons and say they hammered that week, but their playing still demonstrates an unacceptable lack of control, I tell the student they did not hammer correctly that week, or they did not hammer enough. I have found no exceptions to this:  if it isn’t producing results, then it isn’t being used correctly.

Every time a student says that hammering isn’t working despite spending a lot of time with it (or when they think it’s working but I can hear it’s not), it’s because when they do it, they’re not listening closely enough to the exact volume of every note, the exact timing of every note, the exact lifting of every note (clarity), and/or the exact togetherness of notes between the hands.  If any one of these factors is uncontrolled, then hammering will serve little or no benefit.  This can be caused by a simple matter of lack of concentration, or it can be from not slowing down enough – a fast tempo can make it difficult for someone inexperienced at this to monitor/detect problems.  Remember that both hands (not just the hand with the melody) will be played equally loud when hammering, and best to do it with no pedal – mistakes are “nakedly” apparent that way, which is just the way you want them.

Sometimes students do everything right, but they slow down too much.  In that case, all of the challenge of the passagework is gone, and the hammering experience is too drastically different from the performance experience, to the point where they might as well be practicing a different piece altogether.  The hammer tempo needs to be fast enough to be at least a little challenging, but slow enough to allow problems to be identified and corrected.

A pianist must stay relaxed constantly and let their dead arm weight rest on the fingers. It should feel “easy” to play loud.  But this shouldn’t be confused with sluggish movement.  Your fingers will still move quickly both when depressing and when lifting off keys.

How often a pianist hammers a piece might be 10 to 1 (10 times for every one time it is performed) for a piece that sounds “out of control” when performed.  But that ratio might flip once control is harnessed – hammering it only once for every 10 times it’s performed may be all that is necessary.

A piece that has already been learned and severely lacks control will surely take longer than a week to fix.  There are exceptions, such as a student of mine who was working on the Prelude of Bach’s English Suite No. 2 for a competition, and after only a week of hammering, it sounded utterly transformed.  Keep in mind this was an obsessive student who worked hours each day, not to mention the fact that he had actually learned quite a few pieces in the past by watching YouTube videos of musical performances that are generated via MIDI data, scrolling the MIDI data visually in piano roll format.  He was no stranger to the sound of perfect, computer-like musical precision.

The good news is that hammering (even if done wrong) will not actually “harm” a piece in any way.  It may not help it, but it certainly won’t harm it, as long as the pianist continues to also practice each day the musical image of the piece that was carefully constructed by the student and teacher.

A Real Life Example From a Reader

A student pianist e-mailed me in January 2019 and said that he was preparing for a competition. He would be playing Brahms’ Rhapsody in G minor, Op. 79, No. 2, Haydn’s Sonata in E minor, Hob. 34 (I), and Debussy’s Pagodes from Estampes. He said he came across this hammering article as he searched for a way to address his dissatisfaction with his playing. After a few e-mails, he wrote:

In the Debussy piece (Estampes Pagodes), the runs at the end of the piece are really challenging, and when I go at full tempo, I miss some notes, do you have any ideas as to how I can fix this issue?

For those unfamiliar with this piece, the runs at the end are quite difficult and very soft (pianissimo), similar to the passagework one encounters in Ravel’s Jeux d’eau. Pianists and teachers, especially experienced ones not familiar with hammering, might identify this silky, smooth passagework as the prime example of musical material that should not be hammered. After all, hammering is done with both hands equally forte, with no pedal, which makes this section sound almost comically wrong. But, as it turns out, the most dramatically positive results in hammering can be seen precisely in passages like this.  I was so excited to get this question from the student, because I knew exactly where it would inevitably lead. There could only be one ending to this story, and it had to be transformative.

I responded:

Glad to help.

Ghosted notes (missing notes) is an issue of dynamic evenness, which again is one of the many aspects that contribute to the general idea of “control.”  Hammering gives more control.  So as lame as the answer might sound, the answer is the same:  hammer that passagework!  Some might tell you that you should only practice places like that softly, but these types of soft passages actually benefit more from hammering than louder passagework because all of the soft practicing gives the fingers so little tactile feedback.  Feedback is part of how our muscle memory develops.

Three weeks later, I received the following e-mail from him out of the blue:

I’d like to thank you so much for your advice. I won the competition after implementing the hammering technique. Thank you so much.

I congratulated him and his teacher, and for the benefit of my readers, I explicitly asked:

I was also curious if your teacher noticed a difference in your playing before writing me vs. just before (or during) the competition?

He replied:

My teacher did indeed notice the new colours and dexterity I gained in my playing after I found your article, and I think it helped the Debussy piece the most.

Would he have won the competition without hammering? Maybe, maybe not. But speaking as someone who has adjudicated many competitions, I know that displaying control over what we play (rhythmic, dynamic, clarity and togetherness) is one of the biggest advantages any performer can have in a competition.

Objections

Here is an e-mail from an anonymous reader who has experienced the benefits of hammering but who felt the need to hide this practice technique from his teacher:

At the time [as a senior in high school] I was working on passage work and scales as part of an acid bath I was giving myself in improving my articulation, and found through experimentation (maybe frustration one day?!) that playing through these passages very deliberately and forte was really improving my finger work, and that after enough of this work, when brought back to the required softer tone of the passage in question in the repertoire, the notes rippled out and sounded so much more ’rounded.’ At the same time when  asked my chamber music tutor about a particularly knotty passage and how I could practice it, she replied (perhaps in jest) that I should “play it 100 times every day very loudly!” I did, partly to show her I had the discipline to follow her instruction, but found along the way that this made the barrage of semiquavers flow readily and felt easy and as if there was something in reserve. I then read some time later that Horowitz and Richer also practiced very loudly, and perhaps also Josef Lhévinne too, if memory serves. Perhaps this was a Russian school modus operandi. In any case, I did it without my principal teacher knowing as she would have decried the practice as being unmusical, but I knew for me it produced results and indulged in it when at home, and so I was delighted to see someone had written about this procedure, and unabashedly so. So know I know what I was doing back then – hammering!!

Unfortunately, this is not uncommon. Many high-level teachers have a philosophical / emotional objection to hammering that serves no purpose in the realm of good pedagogy. Remember that pedagogy is supposed to be a scientific approach toward teaching. For these teachers, music is art, and no art should be treated like a sport by engaging in physical training. They say that nothing we do to produce this art or emotion can resemble something that doesn’t seem very artistic. Since “playing like a computer” sounds mechanical, they conclude that hammering while practicing will only lead to a performance that sounds mechanical, harsh, etc. This is an unfortunate, extremely simplistic view of how musical and technical development work. The fact is, a pianist can play softer and with more sensitivity when they have complete control over the music, and that’s what hammering produces – control. A pianist is still left to come up with intoxicating musical ideas on their own, but at least hammering makes those ideas more possible to turn into physical reality once those ideas arrive.

When teachers hear a student failing to control passagework, we teachers universally recognize that we are hearing a technical problem, not a musical one. And yet some teachers are caught in a contradiction when they claim that this technical solution shouldn’t be used to correct any problem since it doesn’t sound musical. Shouldn’t a musical problem be addressed musically, and a technical problem technically? The idea that we need to come up with an artistically, aesthetically-pleasing way to solve every pianistic problem we encounter is an arbitrary game that is a huge disservice to the student who cares more about effective solutions than pretty solutions.

Similarly, I’ve also heard a world-renown pianist/professor say that we teachers must absolutely forbid ourselves from ever using words like “hammer,” “slam,” “pound,” etc. when giving students instruction. Again, this is nothing more than a personal/philosophical objection to these words which has no basis in good pedagogy. Show me the study that shows that no pianist who has ever spoken or thought of “hammering” or “slamming” the piano has never once won an international piano competition. Ever since then, I’ve gone out of my way to use words like that when they are musically appropriate, such as with Prokofiev or certain fortissimo/fortississimo sections of other composers, and just as I expected, it hasn’t had any positive or negative effect whatsoever on my students’ playing – it has produced the same effect as if I had said, “Make the piano roar” or “Give it all you got.” Contrary to what these pianists often believe, while audiences can certainly hear the dynamic level of a struck note, they cannot telepathically hear non-musical adjectives that reside in the brains of pianists as they strike notes. I think better advice would have been, “Only use these words when they match the effect you’re trying to produce on the piano.”

Yes, it is true that sometimes certain adjectives can have positive or negative effects on the particular technique a student uses to play a passage.  But this effect cannot be generalized to the entire population.  Even the most extreme adjectives such as “slam” and “pound” will make one student do one thing (play with forearms) while another student does something different (play partially with forearms but mostly with hands dropping from the wrist joint).  One never wants to “slam” at the cadence of a Bach fugue, but to ask people never to use this word – ever – in piano instruction, is nothing more than a cranky and unnecessary overreaction to loud, percussive playing.

The other objection I’ve heard against hammering is a more general objection to the study of technique at the piano. Those who believe that repertoire alone is sufficient for technical development are, from what I have observed so far (without exception), grown up prodigies. They are people who neglected technical exercises as children and developed phenomenal technique anyway. They have tentacles growing directly out of their brains that make the piano keys do anything they want just by willing it to happen, and they assume that when other people play with suffering technique, it’s a musical problem that should be fixed only by creating a more accurate audio image of the music.  (This is a real argument I encountered by a Polish instructor who teaches using the “Chopin Method.”) Their natural technique has made technique invisible to them. So, of course, they will never understand the value of hammering and will consequently make fun of it the same way the prodigious Renaissance man Saint-Saens makes fun of lowly pianists not born with his gifts in Carnival of the Animals (i.e. pianists who practice exercises are apparently chimps, not humans).

References

Doidge, Norman. The Brain That Changes Itself: Stories of Personal Triumph from the Frontiers of Brain Science. New York: Viking, 2007.

(c) 2014 Cerebroom

  1. It might seem that lack of togetherness is the same as rhythmic evenness, but while rhythmic unevenness will certainly cause lack of togetherness (and vise-versa), it still isn’t quite the same thing from the point of view of human perception. The rhythmic interaction between fingers on different arms of our body feels like a different phenomenon to the human mind than the mere sequence of one linear rhythm in one hand. For those who might follow this esoteric analogy, I think of it kind of like how piano technicians hear beats when they’re tuning a string:  a beat is the newly-created frequency of two separate frequencies interacting with each other. Just as a piano technician cannot evaluate or fix beats without the presence of at least two vibrating strings, a pianist also cannot evaluate or fix lack of togetherness by playing hands alone.
  2. Rhythmic unevenness always consists of rushing, not dragging – if rhythm drags, then it’s a dexterity issue, not a control issue.
  3. It would be interesting to know where Mark Westcott learned this technique, where his teacher learned it, and ultimately who the first pianist or keyboardist was to utilize hammering.
  4. In my high school years, I notated around 100 pages of music of Mendelssohn into a computer program called Visual Composer by Adlib. I had to go to a lot of trouble to add dynamics, balance and voicing to the “.rol” files I created, and until I did, the music sounded hammered.

One thought on “Hammering: Gaining and Maintaining Control in Dexterous Piano Playing

Leave a Reply